Shop OBEX P1 Docs P2 Docs Learn Events
Medium current driver — Parallax Forums

Medium current driver

dev/nulldev/null Posts: 381
edited 2009-09-21 15:02 in General Discussion
I'm looking for a transistor to run a medium current DC motor.

I've got the cables, but can't find a proper MOSFET. Anyone?

800px-CERN-cables-p1030764.jpg

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't worry. Be happy

Comments

  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-09-19 03:52
    What kind of currents are you talking about? If your picture is your idea of a medium current, I can't help.

    However, we need to know what kind of voltage you are talking about as well...
  • dev/nulldev/null Posts: 381
    edited 2009-09-19 04:41
    The cables are used to power an LHC unipolar stepper motor with 1600 coils. At 660000 RPM it needs about 1200000 Volts.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Don't worry. Be happy
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2009-09-19 13:03
    ·I think somebody's being facetious (or he's taking the p***.)
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-09-19 14:16
    What would be considered 'HIGH" current I wonder? ... I don't doubt that the 'wire-brick' can handle 12500 Amps, according to the American Wire Gauge standards, a 0.55 inch diameter wire can handle about 500 Amps, so 25 of those wires bundled together works out to 12500 Amps.

    dev/null,

    You need to provide more information on your motor. If you are dealing with 12500 Amps, then I'm sure that a single MOSFET won't cut it for you. The art of combining MOSFETS in parallel to achieve very high current control can be very tricky, everything must be perfectly balanced.

    About the highest current controller of that I am aware of is used in the EV (Electronic Vehicle) industry where 500 Amps is common.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • Agent420Agent420 Posts: 439
    edited 2009-09-19 18:13
    I note those cables are French... must be low resistance devil.gif

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-09-19 18:49
    Ah, come on Agent420 - that was just mean... Even if it were true.
  • dev/nulldev/null Posts: 381
    edited 2009-09-19 22:37
    PJ glad you got it.

    Guys, these cables are used to power a particle accelerator with 1600 magnets at 8 Tesla each! The motor took 7 years to build and cost approx. $1 billion.
    The lower wire is a superconductor (zero resistanse) that handles 12K Amps at around 3 Kelvin.

    en.wikipedia.org/wiki/Large_Hadron_Collider

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Don't worry. Be happy
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2009-09-19 23:21
    dev/null,

    You can't kid a kidder.



    · lol.gif
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-09-20 05:12
    I thought it was powering a 12MV unipolar stepper motor? LOL

    Serioiusly, what are you working on? If you want a MOSFET recommendation, let us know your specs.
  • dev/nulldev/null Posts: 381
    edited 2009-09-20 12:59
    It was a pun Phil, bad hangover humor smile.gif

    Thanks anyway. Actually, when I think about it, I need a P-channel MOSFET with continuous drain current minimum 60 amps.
    I know there is the IRF4905. Are there others?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Don't worry. Be happy
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-09-20 15:45
    Are you sure you need a P-Channel? They are usually more expensive, and generally just not "as good" as N-Channels. You can get High-Side N-Channel drivers which will get the N-Channel's to "act" like a P-Channel in the sense that you can connect a load to the source, and still operate the device.

    Mind posting your schematic?
  • AleAle Posts: 2,363
    edited 2009-09-20 17:39
    If you are serious about those currents... then Semikron has them.

    Have fun

    Ale

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Visit some of my articles at Propeller Wiki:
    MATH on the propeller propeller.wikispaces.com/MATH
    pPropQL: propeller.wikispaces.com/pPropQL
    pPropQL020: propeller.wikispaces.com/pPropQL020
    OMU for the pPropQL/020 propeller.wikispaces.com/OMU
  • dev/nulldev/null Posts: 381
    edited 2009-09-20 18:43
    Phil, it's a simple H-bridge to drive inductive loads. The P-channel FET's are needed on the high side. It looks to me that all Semokrons are N-channel? Any suggestions for high-side N-channel types?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Don't worry. Be happy
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-09-20 20:24
    dev/null,

    Here is something comparable to the IRF4905 that you mentioned.

    www.st.com/stonline/products/literature/ds/8177/stp80pf55.pdf
    It has a lower Rds"ON" than the IRF4905 and can handle 80 Amps.

    It would help though to know what your schematic looks like... not all H-bridges are created equal.
    Also, to know what voltage you are using. There are some MOSFETS that go beyond 100Amp capabilities and have much lower Rds"ON", but at cost of a lower operating voltage than the 55V limitation indicated by the IRF4905 you mentioned.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • dev/nulldev/null Posts: 381
    edited 2009-09-20 22:13
    Here's the circuit. I want to replace the IRF4905's as they behave poorly at 50 amps and above.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Don't worry. Be happy
    Image2.jpg
    789 x 605 - 50K
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2009-09-21 00:00
    I would not think that circuit is suitable for such high currents AND PWM! I have used similar opamps to drive mosfets but considering that these opamps have limited drive capability and the mosfets have such large gate capacitance they will turn on and off slowly. This guarantees the mosfets will get hot and if the frequency is a bit too high then they won't ever be able to turn on fully anyway, certainly not performing like the perfect switches you would like. The 10K gate resistors will only compound this problem. You should also be aware that the gate will breakdown with voltages over 20V or even less so the drive voltage needs to be limited or clamped.

    *Peter*
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-09-21 01:02
    Dev/null, as a general rule, don't EVER use an opamp for frequencies above 1kHz. The OpAmp just can't supply the current to charge the gate fast enough.

    National Semiconductor and TI make some good High-Side mosfet drivers. They are designed to use an N-channel fet in place of the P-channel variety.
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-09-21 03:52
    You also want to avoid high speeds with that circuit because of the way that the NPN's are cross coupled creating an unbalanced hysteresis effect on the inputs.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • dev/nulldev/null Posts: 381
    edited 2009-09-21 14:30
    Hi all, thanks for all the info.

    If I want an inductive driver that handles 60+ amps, what is the best way to go? I do not need PWM in particular, the point is to vary the voltage across the motor coils.
    (And yes, I want to build it myself)

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Don't worry. Be happy
  • AleAle Posts: 2,363
    edited 2009-09-21 15:02
    Why do you want variable voltage across the coils ? For that purpose you use PWM, you do not need variable voltage. Keeping the full voltage keeps the torque at low rpms, something you will lose with low voltage.
    But If you wanted you would have to do a "normal" linear regulator but for that amount of current... no I wouldn't.
    There are MOSFET drivers that take care of keeping just one mosfet in conduction where the other one is in no conduction reducing the losses you will incur when both conduct due to Toff longer than Ton. I'd use one of those, you can also built that yourself. Almost any manufacturer has them: Analog, TI, ST, International Rectifier and so on.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Visit some of my articles at Propeller Wiki:
    MATH on the propeller propeller.wikispaces.com/MATH
    pPropQL: propeller.wikispaces.com/pPropQL
    pPropQL020: propeller.wikispaces.com/pPropQL020
    OMU for the pPropQL/020 propeller.wikispaces.com/OMU
Sign In or Register to comment.